Who Should Attend?
This program is most appropriate for individuals interested in learning about machine learning, with a focus on recent algorithms, like deep learning. You’ll learn the mathematics and statistics at the foundation of modern machine learning and get hands-on training in the latest machine learning software, using Google TensorFlow platform.
You should have a strong background in computing (e.g., Python, Matlab, SAS, etc.; any modern computing language), to be capable of learning how to use and apply modern machine learning software. For participants who also have a strong mathematical and statistical background (strength in calculus and in basic statistics, at the senior undergraduate level), the opportunity to understand the fundamentals of machine learning will be available. Strength in mathematics and statistics is a significant plus; however, it is not required to benefit from the hands-on software portion of the program.
More about this program
Lecturer: David Carlson
Content: Basic concepts in machine learning
- Introduction to model building
- Scaling to “big data” with stochastic gradient descent
- Backpropagation as an efficient computation method
Lecturer: Tim Dunn
Content: Deep convolutional neural networks
- Image analysis
- Image segmentation, object detection and object localization
Lecturer: Lawrence Carin
Content: Reinforcement Learning
- Basic concepts for optimal policies in complex environments
- Q-learning and leveraging deep networks
- Applications of reinforcement learnings
Lecturer: Ricardo Henao
Content: Data synthesis, with an emphasis on images
- Generative adversarial network (GAN)
- Deep networks for GAN
- Learning and applications of GAN
Lecturer: Mohit Bansal
Content: Methods for natural language processing
- Word embeddings
- Recurrent neural networks
- Temporal convolutional neural networks
This five-day program will offer lectures on the mathematics and statistics at the heart of machine learning, plus hands-on training about implementing machine learning tools with the TensorFlow software platform. Each day, material will be discussed at three levels. First, concepts will be presented in an intuitive manner, with light emphasis on the mathematical details. The second portion of each day will then examine the underlying mathematics and statistics of the machine learning algorithms in greater detail. The third portion of each day will focus on software implementation in TensorFlow. Finally, breakout sessions for reviews will be presented in the final hour each day, and there will also be special-topic lectures in the last hour; each student may select from among the parallel activities in the final hour.
Each day will be arranged as follows:
Lecture 1: Mathematically-light introduction to the focus of the day
Lecture 2: Mathematically rigorous discussion of the focus of the day
Software discussion and hands-on training with TensorFlow
Three parallel activities will take place in the last hour each day, and each student may select what is most appropriate for them: (1) In the main lecture hall, there will be a special-topic presentation on an important area of machine learning; (2) there will be a dedicated breakout session for attendees who are medical professionals, to place the earlier lectures in the context of healthcare (Matthew Englehard, MD/PhD, will lead these reviews); and (3) there will be breakout sessions for all other students, to review the earlier lectures that day (led by Lawrence Carin).
The agenda for the special-top lectures (4-5pm) is as follows:
Day 1: The concept of interpretable machine learning (lecturer: Cynthia Rudin)
Day 2: Interpretable machine learning in practice (lecturer: Cynthia Rudin)
Day 3: Applications of machine learning in vision (lecturer: Guillermo Sapiro)
Day 4: Machine learning for face recognition (lecturer: Guillermo Sapiro)
Day 5: Hardware implementations of machine learning (lecturer: Helen Li)
At the end of the program, you should be able to use TensorFlow to implement the latest machine learning methods for analyzing images, video, and natural language (text). For those with a strong mathematical background, the underlying methodology of machine learning will also be covered. You will be given assignments to test your knowledge of the material, so you can get a sense of how well you have absorbed these concepts. Breakout sessions will also be held to offer clarification on concepts and help with hands-on software implementations on provided example datasets.